Incompleteness of an evolving paradigm based on its reliance on poorly elaborated models should not be considered a valid criticism. Progressive elaboration is inherent in development efforts as repeated cycles of planning enable better information, details, estimates, and the overall plan becomes more complete .
Exceeding the scope of good practices however, is a valid target for criticism. Apparent contradictions to minimally elaborated model will usually be examined carefully by model developers. (See Model & Analogy Evaluation)
Revolutionary changes often occur when undirected, diverse approaches to measuring or forecasting produce surprising results. This was recently the case with a model from the UK, which predicted an inverse relationship between surface sea temperature and low cloud cover, providing positive feedback to temperature oscillations. This model predicted more accurately than any of the other 18 simulations reviewed, and it was also closest to the average of the results of the group. Positive feedback mechanisms are destabilizing, and finding them in climate models of current weather generates reasonable concern.
Whether this particular model (or just averaging the others a la "The Wisdom of Crowds") contains processes that meaningfully relate to actual ocean and cloud interactions is not completely certain but, absent other obvious constraints to incorporate into the model, this possibility is the most likely area for productive investigation. If we look at debates over evolution and human caused climate change that are considered settled within the research community, we often see critics claim incompleteness as a defect.
In measuring the maturity and explanatory of a paradigm, we may choose to examine such characteristics of criticism, as these will change over time and in different ways for successful and unsuccessful concepts.
Favorite line of the day: "...conceptual innovation involves changing and creating new constraints", (Nersessian 2009) p.188